123 Main Street, New York, NY 10001

Submit Your BOM

Noise Metrics for INAs: 0.1–10 Hz and Wideband

Noise Metrics for INAs: 0.1–10 Hz and Wideband

← Back to:Instrumentation Amplifiers (INA)

This page shows how to read 0.1–10 Hz peak-to-peak noise and wideband noise density correctly, then turn them into integrated RMS noise and real, bandwidth-defined resolution. It also provides repeatable measurement rules and pass/fail criteria so noise results stay consistent across benches, boards, and production.

What “Noise Metrics” Really Mean in an INA Datasheet

INA noise is typically reported using three different lenses: 0.1–10 Hz peak-to-peak, noise density (nV/√Hz), and integrated RMS noise. These are not interchangeable “quality scores.” Each metric answers a different system question, and each becomes meaningful only when paired with a bandwidth, a measurement window, and a clear input-referred reference point.

Key takeaway
0.1–10 Hz p-p describes slow, low-frequency variation that limits DC / quasi-DC resolution. nV/√Hz describes the wideband floor that integrates into RMS over a defined bandwidth. Integrated RMS is only valid when the filter / window (and thus noise bandwidth) is explicitly stated.

The 3 core metrics and what each one answers

0.1–10 Hz noise (peak-to-peak)
  • Answers: “How much slow variation appears in a low-frequency window?”
  • Best for: DC / quasi-DC sensor resolution limits.
  • Pitfall: p-p depends strongly on observation time and windowing; do not compare p-p numbers without matching conditions.
Noise density (nV/√Hz, wideband)
  • Answers: “What is the noise floor per √Hz in the frequency region of interest?”
  • Best for: Computing RMS noise over a defined bandwidth (via integration / ENBW).
  • Pitfall: A single headline number can hide 1/f rise or spurs; use the curve when available.
Integrated noise (RMS over BW / digital window)
  • Answers: “What is the final RMS noise inside a stated bandwidth / filter / window?”
  • Best for: Mapping to resolution, minimum detectable change, and production limits.
  • Pitfall: RMS without bandwidth is not actionable; RMS must be tied to ENBW or an explicit digital window.

Three questions to pick the right metric first

  1. What is the effective bandwidth or observation window? (Noise becomes a number only after bandwidth/window is fixed.)
  2. Is the target dominated by sub-10 Hz behavior or by in-band RMS? (Low-frequency limits vs wideband floor are different problems.)
  3. Is the goal an RMS number or a real “minimum measurable change”? (Resolution mapping requires input-referred RMS + sensor gain chain.)

Boundary rule: low-frequency noise is not drift

A low-frequency noise plot or a 0.1–10 Hz p-p number can be inflated by slow temperature gradients, bias/leakage changes, or recovery from overload. Treat drift and bias-driven slow errors as separate budget owners; keep noise metrics tied to a defined window and a stable thermal condition.

Noise Map (how the metrics relate)

Use noise density to compute integrated RMS over the system’s noise bandwidth (ENBW). Treat RMS ↔ p-p conversion as condition-dependent (window/assumptions), not a fixed constant.

INA Noise Metrics Map Three metrics and their relationship: 0.1–10 Hz peak-to-peak, noise density nV per root Hz, and integrated RMS over bandwidth. Noise metrics: choose → compute → verify 0.1–10 Hz noise (p-p) window-limited Noise density nV/√Hz use curve Integrated noise (RMS) BW / ENBW Use density to compute RMS ENBW / BW Integrated RMS RMS ↔ p-p depends on window & assumptions Do not mix drift

Noise Taxonomy for INAs: 1/f, White, and “Artifacts”

Before converting any datasheet number into “resolution,” it is critical to classify what dominates the spectrum. INA noise is typically a combination of 1/f noise (low-frequency rise), a white-noise floor (flat region), and artifacts (spurs or shaped bumps caused by switching, interference, or measurement chain effects). Correct classification prevents false conclusions and avoids budgets that cannot be met in hardware.

Engineering meaning
1/f region limits DC / sub-10 Hz stability, white floor sets in-band RMS after integration, and artifacts must be treated as separate limits (spur masks), not averaged into noise density.

How to read a PSD plot (fast rules)

  • Low-frequency slope: rising trend indicates 1/f contribution; it dominates long windows.
  • Flat platform: white-noise floor; integrate this region to estimate in-band RMS.
  • Narrow peaks: spurs (mains, switching ripple, coupling); handle with a mask/limit, not with RMS averaging.
  • Window sensitivity: changing FFT window, sampling rate, or bandwidth changes the displayed floor via ENBW and leakage.

What the 1/f corner changes in real systems

The “corner” is the frequency where 1/f rise meets the white floor. If the effective bandwidth or observation window leans into the 1/f region, low-frequency metrics (including 0.1–10 Hz p-p) will dominate the perceived resolution. If the bandwidth is well above the corner and the front-end remains linear, the white floor dominates the integrated RMS.

Decision rule
Sub-10 Hz / long windows → prioritize 1/f behavior and low-frequency limits. Wideband / dynamic windows → prioritize the flat density floor and correct integration bandwidth (ENBW).

Artifacts that often masquerade as “noise”

Ripple spurs
Fixed-frequency peaks (often from switching/chopping or coupling). Quick recognition: peak stays at the same frequency across bandwidth changes.
Shaped bumps
Spectrum looks “lifted” or “curved” due to folding/leakage/measurement chain. Quick recognition: shape changes with sampling, window, or anti-alias settings.
Mains / RFI coupling
Narrow peaks (50/60 Hz and harmonics) or clusters. Quick recognition: sensitivity to cable routing, shielding, or probing.

How to budget after classification

Split the “noise” budget into three owners to keep verification unambiguous:

  • LF (1/f) owner: use low-frequency windows (e.g., 0.1–10 Hz p-p) under controlled thermal conditions.
  • WB (white) owner: integrate density over ENBW to compute RMS in the specified bandwidth.
  • Artifacts owner: define spur masks / limits (peak X relative to full-scale or target noise budget), rather than hiding peaks inside RMS.
PSD Zones (1/f + white + spur)

A PSD plot can be read as three regions: 1/f rise, flat white floor, and narrow spurs. Correctly identifying the dominant region prevents mixing interference and drift-like behavior into “noise” claims.

INA Noise PSD Zones Log-frequency PSD sketch showing 1/f region, white noise floor, corner point, and a ripple spur peak. PSD zoning: identify what dominates Frequency (log) Noise density corner 1/f white floor ripple spur Treat spurs as separate limits; integrate only the noise floor over ENBW

Understanding 0.1–10 Hz Noise (Peak-to-Peak): What It Captures and What It Hides

The 0.1–10 Hz peak-to-peak number is a windowed, low-frequency view of input-referred variation. It is designed to reflect how “stable” a DC or slow sensor reading looks over time. Unlike RMS metrics, peak-to-peak is not a stationary statistic—it changes with observation time, filtering method, and sample count. A meaningful p-p requirement must always be tied to a stated measurement window and band-pass method.

Key takeaway
0.1–10 Hz p-p captures slow in-window variation that limits DC / quasi-DC resolution. It hides the fact that peak-to-peak grows with longer observation time and depends on the band-pass implementation. Do not compare p-p numbers unless the window, band-pass, and conditions match.

What 0.1–10 Hz p-p is actually measuring

  • Step 1: Record a time series for an observation time Tobs.
  • Step 2: Apply a 0.1–10 Hz band-pass (analog or digital).
  • Step 3: Compute peak-to-peak as max − min over the filtered record.
  • Interpretation: The result is a window-limited “how much it moves” metric, not a bandwidth-agnostic noise floor.

Why peak-to-peak is statistically unstable

Peak-to-peak is an extreme-value statistic. With more samples (longer Tobs or higher sample rate), the probability of observing a larger excursion increases. Band-pass details (transition band, group delay, digital windowing) also change the “shape” of the filtered waveform and therefore the peaks that are seen.

Practical rule
If two results use different Tobs or different band-pass implementations, treat them as different tests, not directly comparable numbers.

Common mistakes that create “inconsistent” results

  • Comparing p-p to RMS directly: p-p depends on window and assumptions; RMS depends on bandwidth/ENBW.
  • Comparing different record lengths: longer Tobs usually increases p-p even when the device is unchanged.
  • Including slow components: thermal gradients, bias/leakage changes, or recovery effects can inflate the in-window variation.
  • Including narrowband interference: mains or coupling spurs are not “noise floor” and should be handled as separate limits.

How to specify a comparable 0.1–10 Hz p-p requirement

A production-ready p-p requirement must include the minimum reporting fields below. Without them, pass/fail cannot be reproduced.

Minimum reporting fields
  • Band-pass: 0.1–10 Hz (method: analog/digital)
  • Observation time: Tobs = ___
  • Sampling rate (if digital): fs = ___
  • Input condition: shorted / stated source impedance
  • Thermal condition: steady-state (no airflow steps)
Pass criteria template
0.1–10 Hz noise (p-p), input-referred
Tobs = ___ s, band-pass = ___
< X µVpp
Time series + window (how p-p is formed)

Peak-to-peak is taken after the 0.1–10 Hz band-pass and within the chosen observation window. Longer windows increase the chance of larger extremes.

0.1–10 Hz Peak-to-Peak: Windowed Measurement Time series shows slow component and noise, band-pass 0.1–10 Hz, and peak-to-peak measured over observation time Tobs. 0.1–10 Hz p-p is a windowed statistic Raw time series slow component Tobs 0.1–10 Hz band-pass Band-passed result max / min p-p

Wideband Noise Density (nV/√Hz): How to Read and Use It

Noise density expresses the RMS noise per √Hz at a given frequency. It becomes a system number only after integration over the effective noise bandwidth set by the analog filter or the digital window. Because INA density is often frequency-dependent (1/f rise at low frequency and a flat white floor at higher frequency), a single headline value can be misleading unless it matches the frequency region of interest.

Key takeaway
Treat nV/√Hz as a curve, not a single score. Read the curve in the frequency region that matters, then integrate over ENBW to obtain RMS. Add source resistance noise in the same density domain before integration.

What nV/√Hz means (system translation)

  • Density domain: noise is expressed per √Hz at a specific frequency.
  • RMS domain: RMS noise is obtained by integrating density over ENBW.
  • Design implication: increasing bandwidth increases integrated RMS even if the density curve is unchanged.

How to read the curve (three read points)

Use the curve to identify whether the operating band sits in the 1/f rise or in the white floor. A single headline density value is only valid when it is read in the same region used for integration.

  • 10 Hz: checks low-frequency rise proximity.
  • White floor: the flat region used for most wideband RMS calculations.
  • 1 kHz: a practical in-band anchor for many conditioned sensors.

Source resistance noise (combine in density first)

A sensor or source resistance contributes its own noise in the same nV/√Hz domain. Combine independent density contributors using an RMS-in-density approach before integrating:

Combine (input-referred density)
e_total ≈ √( e_INA² + e_source² + e_chain² )
Then integrate e_total over ENBW to obtain RMS.

Production-ready reporting fields for density

Minimum reporting fields
  • Density point: en @ f = ___ nV/√Hz
  • Gain and input reference: input-referred statement
  • Source impedance stated (shorted or Rs = ___)
  • Bandwidth / filter context (for later integration)
  • Spur handling: spur mask / limits kept separate
Pass criteria template
Noise density (white floor region)
en ≤ X nV/√Hz @ f = ___
Conditions: Rs = ___, gain = ___
Density curve + read points (use curve, not a headline)

A typical INA density curve rises at low frequency (1/f) and flattens into a white floor. Read density at representative points, then integrate over ENBW to obtain RMS.

Noise Density Curve Read Points Log-frequency noise density sketch showing 1/f rise, white noise floor, and read points at 10 Hz and 1 kHz. Noise density: use the curve Frequency (log) Noise density (nV/√Hz) 1/f rise white floor 10 Hz 1 kHz read here Use curve → combine densities → integrate over ENBW → RMS

Converting Noise Density to Integrated RMS Noise

Integrated RMS noise is the input-referred RMS voltage that results after the noise density curve is shaped by a measurement bandwidth or a filter/window. This number is only meaningful when the noise bandwidth is stated. In practice, the fastest path is to combine density contributors in the nV/√Hz domain, use ENBW to represent the filter/window, then integrate to obtain Vrms and translate it through gain to the output domain.

Rule that prevents wrong calculations
BW here is noise bandwidth, not “-3 dB bandwidth.” If a filter/window is present, use ENBW as the bandwidth that noise “sees.” Without en(f) (or a stated white-floor value) and a stated BW/ENBW, an RMS number is not reproducible.

Model 1: Pure white noise (fast engineering estimate)

Use when the operating band sits in the white floor region of the density curve (away from the 1/f rise). Read a representative en_white, then apply the noise bandwidth.

Steps
  • Read en_white from the flat region of en(f).
  • Use BW_noise or ENBW as the bandwidth.
  • Compute Vrms_in ≈ en_white · √(BW_noise).
  • Translate: Vrms_out = Vrms_in · Gain.

Model 2: 1/f + white (corner-based, piecewise workflow)

Use when low-frequency rise contributes to the band. Treat the curve as two regions separated by a corner frequency where the 1/f rise meets the white floor. This avoids academic math while preserving the correct dependency on the operating limits.

Steps
  • Find f_c from en(f): where the 1/f rise meets the flat floor.
  • Define band limits f_L and f_H from the measurement window/filter.
  • Estimate low-frequency contribution using the curve segment over [f_L, f_c].
  • Estimate white-floor contribution using en_white over [f_c, f_H].
  • Combine the two contributions in RMS (do not add amplitudes).
Important boundary
If f_L is set by a digital window or detrending step, it must be stated. Changing the window changes the low-frequency contribution and the final Vrms.

Model 3: With a filter/window (use ENBW, not -3 dB)

Analog anti-alias filters, RC networks, and digital averaging windows shape noise. Represent that shaping using ENBW (effective noise bandwidth). Once ENBW is known, apply Model 1 or Model 2 using BW_noise = ENBW.

Minimum fields
  • Filter/window type and settings
  • ENBW value (from datasheet, sim, or measurement)
  • Band limits f_L, f_H (if applicable)
  • Gain and reference point (input-referred statement)
Result structure
Vrms_in (input-referred)
Vrms_in computed with BW_noise = ENBW
Vrms_out (output domain)
Vrms_out = Vrms_in · Gain

Sanity checks (catch wrong bandwidth and spur mixing)

BW scaling
White-noise dominated Vrms should scale approximately with √(BW_noise). If it does not, the operating band is not in the white floor or the window is changing f_L.
Source impedance sensitivity
Shorted input vs stated source resistance should change the total density as expected. If not, narrowband interference or leakage paths are dominating.
Spur separation
Do not fold mains or switching spurs into “noise floor.” Treat spurs with a separate mask/limit and keep Vrms reserved for broadband noise.
Integration workflow (en(f) → ENBW → Vrms)

Combine independent contributors in the density domain, apply the filter/window through ENBW, integrate to input-referred Vrms, then translate through gain.

Noise Density to Integrated RMS Using ENBW Block diagram showing en(f) plus source noise, filter/window H(f), ENBW and integration producing Vrms_in and Vrms_out through gain. Integration flow: density → ENBW → Vrms en(f) source noise (Rs) H(f) filter / window ENBW noise BW ≠ -3 dB Integrate Vrms_in × Gain → Vrms_out Combine in density √(e1² + e2² + …) before integration

RMS ↔ Peak-to-Peak: When the Conversion is Valid (and When It’s Not)

The ratio k = Vpp / Vrms is not a universal constant. Peak-to-peak is an extreme-value statistic that depends on observation time, bandwidth/window, and whether the record contains slow components that are not part of broadband noise. A conversion is only meaningful when the signal is approximately Gaussian and the measurement window and bandwidth are fixed and explicitly stated.

Rule that prevents “magic k” misuse
Use a k range tied to a fixed window, not a single number. If Tobs changes, peak-to-peak usually increases and k increases. If slow components are present, the conversion no longer represents broadband noise.

When a conversion is valid

  • Noise is close to Gaussian within the stated band.
  • Window and bandwidth are fixed: same Tobs and same band-pass/ENBW.
  • Measurement conditions are steady: no step changes in temperature, recovery, or bias state.
  • Spurs are treated separately: the record is not dominated by narrowband interference.

Why “one k” fails across different windows

Peak-to-peak grows with longer observation because more samples increase the chance of seeing a larger extreme. RMS does not grow the same way when bandwidth is fixed. Therefore, k changes with Tobs even when the underlying density curve is unchanged.

Spec writing pattern
Always state: Tobs + band-pass (or ENBW) for both Vpp and Vrms. Avoid “Vpp converted to Vrms” unless the test conditions are identical.

How to self-calibrate k in a real system (recommended)

  1. Fix the measurement definition: Tobs and the band-pass/ENBW.
  2. Fix the input condition: shorted input or stated source impedance.
  3. Capture one record and compute Vrms and Vpp on the same window.
  4. Repeat across multiple records to obtain a k distribution.
  5. Pick a conservative k rule that matches the acceptance philosophy (range-based).
  6. Store k together with the test definition as part of the production spec.

Boundary reminder: do not mix slow components into noise conversion

If the record contains slow components (thermal gradients, bias/leakage changes, overload recovery), Vpp inflates and k no longer represents broadband noise. Treat slow error owners under DC accuracy and keep noise conversion limited to a defined band/window.

Practical alternative
Use Vrms for resolution/threshold mapping, and use Vpp for “worst observed movement,” each with a fixed definition.
k vs observation time (why p-p grows with longer windows)

For a fixed bandwidth, longer observation usually increases the extreme values seen in a record. That pushes peak-to-peak upward and increases k.

k = Vpp/Vrms Increases With Observation Time Simple trend chart showing k rising as observation time increases, highlighting why a single conversion factor is not universal. Conversion factor is window-dependent Tobs k = Vpp / Vrms short window longer observation Longer observation → larger p-p → larger k Use a k range tied to a fixed window and bandwidth

Converting Noise Density to Integrated RMS Noise

Integrated RMS noise is the input-referred RMS voltage that results after the noise density curve is shaped by a measurement bandwidth or a filter/window. This number is only meaningful when the noise bandwidth is stated. In practice, the fastest path is to combine density contributors in the nV/√Hz domain, use ENBW to represent the filter/window, then integrate to obtain Vrms and translate it through gain to the output domain.

Rule that prevents wrong calculations
BW here is noise bandwidth, not “-3 dB bandwidth.” If a filter/window is present, use ENBW as the bandwidth that noise “sees.” Without en(f) (or a stated white-floor value) and a stated BW/ENBW, an RMS number is not reproducible.

Model 1: Pure white noise (fast engineering estimate)

Use when the operating band sits in the white floor region of the density curve (away from the 1/f rise). Read a representative en_white, then apply the noise bandwidth.

Steps
  • Read en_white from the flat region of en(f).
  • Use BW_noise or ENBW as the bandwidth.
  • Compute Vrms_in ≈ en_white · √(BW_noise).
  • Translate: Vrms_out = Vrms_in · Gain.

Model 2: 1/f + white (corner-based, piecewise workflow)

Use when low-frequency rise contributes to the band. Treat the curve as two regions separated by a corner frequency where the 1/f rise meets the white floor. This avoids academic math while preserving the correct dependency on the operating limits.

Steps
  • Find f_c from en(f): where the 1/f rise meets the flat floor.
  • Define band limits f_L and f_H from the measurement window/filter.
  • Estimate low-frequency contribution using the curve segment over [f_L, f_c].
  • Estimate white-floor contribution using en_white over [f_c, f_H].
  • Combine the two contributions in RMS (do not add amplitudes).
Important boundary
If f_L is set by a digital window or detrending step, it must be stated. Changing the window changes the low-frequency contribution and the final Vrms.

Model 3: With a filter/window (use ENBW, not -3 dB)

Analog anti-alias filters, RC networks, and digital averaging windows shape noise. Represent that shaping using ENBW (effective noise bandwidth). Once ENBW is known, apply Model 1 or Model 2 using BW_noise = ENBW.

Minimum fields
  • Filter/window type and settings
  • ENBW value (from datasheet, sim, or measurement)
  • Band limits f_L, f_H (if applicable)
  • Gain and reference point (input-referred statement)
Result structure
Vrms_in (input-referred)
Vrms_in computed with BW_noise = ENBW
Vrms_out (output domain)
Vrms_out = Vrms_in · Gain

Sanity checks (catch wrong bandwidth and spur mixing)

BW scaling
White-noise dominated Vrms should scale approximately with √(BW_noise). If it does not, the operating band is not in the white floor or the window is changing f_L.
Source impedance sensitivity
Shorted input vs stated source resistance should change the total density as expected. If not, narrowband interference or leakage paths are dominating.
Spur separation
Do not fold mains or switching spurs into “noise floor.” Treat spurs with a separate mask/limit and keep Vrms reserved for broadband noise.
Integration workflow (en(f) → ENBW → Vrms)

Combine independent contributors in the density domain, apply the filter/window through ENBW, integrate to input-referred Vrms, then translate through gain.

Noise Density to Integrated RMS Using ENBW Block diagram showing en(f) plus source noise, filter/window H(f), ENBW and integration producing Vrms_in and Vrms_out through gain. Integration flow: density → ENBW → Vrms en(f) source noise (Rs) H(f) filter / window ENBW noise BW ≠ -3 dB Integrate Vrms_in × Gain → Vrms_out Combine in density √(e1² + e2² + …) before integration

RMS ↔ Peak-to-Peak: When the Conversion is Valid (and When It’s Not)

The ratio k = Vpp / Vrms is not a universal constant. Peak-to-peak is an extreme-value statistic that depends on observation time, bandwidth/window, and whether the record contains slow components that are not part of broadband noise. A conversion is only meaningful when the signal is approximately Gaussian and the measurement window and bandwidth are fixed and explicitly stated.

Rule that prevents “magic k” misuse
Use a k range tied to a fixed window, not a single number. If Tobs changes, peak-to-peak usually increases and k increases. If slow components are present, the conversion no longer represents broadband noise.

When a conversion is valid

  • Noise is close to Gaussian within the stated band.
  • Window and bandwidth are fixed: same Tobs and same band-pass/ENBW.
  • Measurement conditions are steady: no step changes in temperature, recovery, or bias state.
  • Spurs are treated separately: the record is not dominated by narrowband interference.

Why “one k” fails across different windows

Peak-to-peak grows with longer observation because more samples increase the chance of seeing a larger extreme. RMS does not grow the same way when bandwidth is fixed. Therefore, k changes with Tobs even when the underlying density curve is unchanged.

Spec writing pattern
Always state: Tobs + band-pass (or ENBW) for both Vpp and Vrms. Avoid “Vpp converted to Vrms” unless the test conditions are identical.

How to self-calibrate k in a real system (recommended)

  1. Fix the measurement definition: Tobs and the band-pass/ENBW.
  2. Fix the input condition: shorted input or stated source impedance.
  3. Capture one record and compute Vrms and Vpp on the same window.
  4. Repeat across multiple records to obtain a k distribution.
  5. Pick a conservative k rule that matches the acceptance philosophy (range-based).
  6. Store k together with the test definition as part of the production spec.

Boundary reminder: do not mix slow components into noise conversion

If the record contains slow components (thermal gradients, bias/leakage changes, overload recovery), Vpp inflates and k no longer represents broadband noise. Treat slow error owners under DC accuracy and keep noise conversion limited to a defined band/window.

Practical alternative
Use Vrms for resolution/threshold mapping, and use Vpp for “worst observed movement,” each with a fixed definition.
k vs observation time (why p-p grows with longer windows)

For a fixed bandwidth, longer observation usually increases the extreme values seen in a record. That pushes peak-to-peak upward and increases k.

k = Vpp/Vrms Increases With Observation Time Simple trend chart showing k rising as observation time increases, highlighting why a single conversion factor is not universal. Conversion factor is window-dependent Tobs k = Vpp / Vrms short window longer observation Longer observation → larger p-p → larger k Use a k range tied to a fixed window and bandwidth

Mapping Noise to Real Sensor Resolution and Bandwidth

Noise metrics become actionable only after converting them into a minimum detectable change under a stated bandwidth and observation definition. Start with input-referred integrated noise (Vrms_in over BW/ENBW), translate it through the signal chain and sensor sensitivity, then express the result as Δsensor_min or an LSB-equivalent value. Bandwidth is the main lever: for broadband noise, larger noise bandwidth increases Vrms and reduces achievable resolution.

One-line engineering rule
BW/ENBW ↑ → Vrms ↑ → resolution ↓. Report resolution only together with the noise bandwidth (or ENBW) and the observation definition.

Step 1: Use a single reference domain (input-referred)

Resolution mapping is most robust when all noise is expressed as input-referred Vrms. If a measurement is made at the output or ADC input, translate back using the stated gain. Keep the bandwidth definition attached to the number.

Required fields
  • Vrms_in over BW/ENBW
  • Total gain to the measurement point
  • Sensor sensitivity S (V per unit or equivalent)
  • Observation definition (window / method)

Step 2: Convert Vrms into a minimum detectable input change

A detection threshold must be defined (decision rule, averaging, and bandwidth). Use a simple template that preserves the dependency on bandwidth without injecting application-specific constants.

Template (input domain)
ΔV_in_min ≈ K · Vrms_in
K = threshold factor (method-dependent)
Template (sensor domain)
Δsensor_min ≈ ΔV_in_min / (S · G_path)
S = sensor sensitivity, G_path = gain mapping

Step 3: Optional LSB-equivalent mapping (measurement chain)

When a digital limit is needed, express the integrated noise at the converter input as a fraction of a code step. Keep the mapping generic and avoid mixing it with converter architecture details.

Template
LSB_equiv ≈ Vrms_at_ADC / V_LSB
V_LSB depends on the chosen coding range

Bandwidth–resolution tradeoff (and a fast validation)

For broadband noise, integrated RMS increases as noise bandwidth increases. That reduces achievable resolution unless averaging or bandwidth limiting is applied. Validate the noise owner by changing BW/ENBW and checking whether Vrms follows the expected scaling.

Engineering conclusion
BW/ENBW ↑ → Vrms ↑ → resolution ↓
Quick check
Keep input condition fixed, change BW/ENBW, and compare Vrms. If Vrms does not follow the expected trend, narrowband spurs or low-frequency components are likely dominating.

Spec writing pattern (make resolution reproducible)

Minimum reporting fields
  • Δsensor_min reported with BW/ENBW = ___
  • Observation definition: window / method = ___
  • Input condition: source impedance / shorted = ___
  • Gain mapping and sensitivity fields stated
Noise → resolution workflow (template fields)

Use a generic chain: sensitivity and gain translate integrated input noise into the minimum detectable sensor change. Bandwidth affects Vrms directly and therefore sets the resolution limit.

Noise to Resolution Mapping Template Block chain showing sensor sensitivity, INA gain, Vrms_in over BW or ENBW, threshold mapping to Delta V and Delta sensor minimum, with a side arrow showing BW increases Vrms and reduces resolution. Template mapping: Vrms → Δsensor S sensitivity G INA gain Vrms_in BW / ENBW ΔV_in_min threshold Δsensor_min result BW ↑ → Vrms ↑ resolution ↓ report BW/ENBW

Chopper / Zero-Drift INAs: Noise Ripple, Folding, and How to Budget It

Zero-drift and chopper INAs often deliver excellent low-frequency stability, but they can introduce structured artifacts that do not behave like random broadband noise. Typical signatures include a fixed spur in the spectrum or a periodic ripple in the time domain. A production-ready noise budget must separate broadband Vrms from spur/ripple limits and evaluate whether the measurement bandwidth window covers the artifact.

Budgeting rule
Broadband noise is integrated to Vrms. Spurs/ripple are handled with a separate spur mask or peak limit. Do not fold a fixed spur into Vrms and call it “noise floor.”

Noise viewpoint (benefit vs cost)

  • Benefit: improved low-frequency stability and reduced low-frequency noise.
  • Cost: ripple/spur signatures that can land inside the measurement band.
  • Risk: artifacts can dominate “resolution” even when the density floor is low.

Separate two acceptance limits

Broadband (noise floor)
Use en(f) and integrate over BW/ENBW to obtain Vrms. This maps to resolution through H2-7.
Structured artifact (spur/ripple)
Treat as a separate limit using a spur mask or a ripple_pp limit over a stated band/window.

Bandwidth window decision (in-band vs out-of-band)

The artifact becomes a resolution limiter when the measurement BW window covers it. If the window excludes the spur, verify that the filter/window provides enough attenuation and that the spur does not fold back through sampling or processing.

  • In-band spur: treat as a ripple/periodic error; set a peak limit.
  • Out-of-band spur: require attenuation margin; verify with spectrum.

Handling principles (keep topology on the AAF page)

Low-pass shaping
Push spur energy out of the measurement band with bandwidth control and verified attenuation margin.
Synchronous sampling
Align acquisition/processing so the artifact becomes predictable and can be suppressed by the chosen window strategy.
Digital filtering
Use a defined bandwidth window (ENBW) or a notch strategy and verify spur placement relative to the window.

Measurable pass criteria (separate Vrms and spur limits)

Broadband
Vrms_in (BW/ENBW = ___) < X
Spur / ripple
spur amplitude below mask OR ripple_pp < X
Spur vs BW window (budget the artifact correctly)

A fixed spur is a structured signature. The key question is whether the measurement bandwidth window covers it. In-band spurs must be handled as ripple or spur limits; out-of-band spurs require verified attenuation margin.

Chopper Spur and Measurement Bandwidth Window Spectrum sketch showing broadband noise floor, a prominent spur, and two bandwidth windows indicating in-band and out-of-band cases. Artifact budgeting depends on BW window Frequency Amplitude noise floor chopper spur in-band out-of-band Spur: separate limit • Vrms: integrate floor over BW/ENBW

Measurement Setups: How to Measure 0.1–10 Hz and Wideband Noise Correctly

Accurate noise measurement depends on a fixed definition: bandwidth (BW/ENBW), observation time, input condition, and processing method. Low-frequency noise is dominated by test environment and stability (shielding, thermal equilibrium, and long observation), while wideband noise demands bandwidth control, FFT/ENBW reporting, and a quantified instrument noise floor. Without measurement-chain self-noise characterization, results cannot be compared to datasheet numbers.

Minimum definition (do not omit)
Always report: BW/ENBW, Tobs, input condition (shorted or stated source impedance), and processing method (band-pass/FFT window and correction). If any of these changes, the reported “noise” changes.

Low-frequency (0.1–10 Hz) setup: stabilize first, then measure

  • Shielding: use an enclosure and keep cables fixed.
  • Thermal equilibrium: wait for steady conditions; avoid airflow changes.
  • Tobs: state observation time; peak-to-peak depends on it.
  • Band-pass definition: specify analog or digital implementation.
  • Input condition: report shorted input vs stated source impedance.

Wideband setup: control bandwidth, then report FFT with ENBW

  • Front-end BW: define the measurement bandwidth or anti-alias limit.
  • Instrument floor: quantify ADC/FFT self-noise before trusting results.
  • FFT settings: report window type and correction approach.
  • ENBW: use ENBW/RBW consistent reporting for comparability.

Quantify measurement-chain self-noise (required)

If the measurement chain is not substantially quieter than the DUT, the reported noise is dominated by the instrument and the conclusion is not actionable.

  • Measure input short baseline (chain floor).
  • Measure with a known source (sanity verification).
  • Change gain and verify noise scaling trend.
  • Use RMS power separation only when chain floor is well below the measurement.

Fast sanity checks (detect definition drift)

BW/ENBW change
Vrms should move with bandwidth trend; if not, spurs or low-frequency components are likely dominating.
Tobs change
0.1–10 Hz peak-to-peak usually increases with longer observation.
Shorted input
If shorting does not reduce the result, the chain floor or coupling dominates.

Reporting template (copy-ready fields)

0.1–10 Hz p-p: ___ (Tobs=___, band-pass=___, fs=___, input=___, shield=___)
Wideband Vrms/PSD: ___ (BW/ENBW=___, AAF=___, FFT=___, window=___, floor=___)
Measurement chain (define, limit, correct, then report)

A stable chain separates DUT noise from instrument floor and ties every number to bandwidth and observation definitions.

Noise Measurement Chain: DUT to FFT Block diagram showing DUT(INA), bandwidth limiting/anti-aliasing, ADC, FFT/window, and metrics, with side tags for shielding, thermal equilibrium, instrument floor, and Tobs/ENBW. Define BW/ENBW + Tobs + floor DUT INA BW limit AA / ENBW report ADC capture FFT window ENBW Shielding Thermal equilibrium Cable fixed Instrument floor quantify Report: BW/ENBW + Tobs method + input condition 0.1–10 Hz p-p PSD / Vrms floor + correction

Common Traps: When “Noise” Is Actually Drift, Hum, Leakage, or Saturation Recovery

A large fraction of “noise problems” are not random broadband noise. The fastest way to converge is to classify the signature and apply a short diagnostic test. The traps below provide a quick check, a bounded fix action, and a pass-criteria placeholder that keeps noise reporting reproducible without expanding into layout or EMI theory.

Classification rule
If the signature is narrowband, treat it as a spur (mask/limit). If it is slow and environment-sensitive, treat it as a low-frequency component that contaminates the noise metric.

Trap A: 50/60 Hz hum (and harmonics)

Quick check
FFT shows a fixed peak at 50/60 Hz and harmonics; shorted input test indicates whether coupling is external or input-driven.
Fix boundary
Treat as spur; apply a spur mask and verify shielding/grounding without folding it into Vrms.
Pass criteria
hum spur < X (relative to floor) OR below spur mask

Trap B: thermal gradients / airflow sensitivity

Quick check
Blocking airflow or adding a cover changes low-frequency energy; repeating after stabilization improves repeatability.
Fix boundary
Stabilize environment and routing; keep the measurement definition fixed. Detailed thermal modeling belongs to the DC-accuracy path.
Pass criteria
Δreading over Tobs < X AND stable within X after settling

Trap C: leakage / contamination / bias path issues

Quick check
Random jumps correlate with humidity or touch; cleaning/drying changes behavior; shorted input does not eliminate events.
Fix boundary
Control leakage paths and cleanliness; layout-level guard and routing details belong to the layout checklist.
Pass criteria
jump events per minute < X OR leakage-induced step < X

Trap D: saturation / overload recovery

Quick check
After a step or mild overload, low-frequency energy stays elevated and decays slowly; repeating with different amplitudes changes the tail.
Fix boundary
Classify as dynamic behavior; avoid saturating conditions and verify recovery under the real stimulus profile.
Pass criteria
recovery time < X AND Vrms returns within X after event
Trap classification board (quick checks first)

Use a fast classification pass to avoid integrating spurs or slow components into a broadband noise metric.

Noise Trap Board: Quick Classification Four-card board listing Hum, Thermal, Leakage, and Recovery traps with minimal signature and quick check tags. Classify the signature before computing Vrms Hum spur FFT peak Thermal slow wander block air Leakage jumps clean & dry Recovery tail step & observe Narrowband → spur mask • Slow components → stabilize environment

Engineering Checklist: Noise Budgeting, Verification, and Production Screens

This checklist turns noise metrics into a reusable workflow: define a minimal noise budget, verify with reproducible measurements, then deploy production screens that are short-time, automatable, and traceable. The scope is noise-only (bandwidth/ENBW, observation time, floor, spurs, repeatability) and does not mix in CMRR, offset/drift, or protection topics.

Noise-only rule
Any reported noise number must carry BW/ENBW, Tobs, input condition, and a floor plan. Without these fields, comparisons across boards, teams, and lots are not meaningful.

Design checklist (minimal noise budget fields)

Budget inputs
  • en(f) curve source and region of use
  • 0.1–10 Hz p-p and stated conditions
  • Noise bandwidth: BW/ENBW (not -3 dB BW)
Mapping fields
  • INA gain and tolerance
  • Sensor sensitivity S (placeholder field)
  • Target resolution: Δsensor or LSB-equivalent
Margin and owner
  • Margin placeholder (X% or X dB)
  • Noise owner assignment (DUT / ADC / BW / environment)
  • Verification plan tied to budget numbers

Verification checklist (five must-measure items)

1) 0.1–10 Hz p-p
Report Tobs and band-pass definition; keep shielding and thermal conditions fixed.
Limit: Vpp_0.1–10Hz < X
2) Wideband density + integrated Vrms
Define front-end BW/AAF; report FFT window and ENBW correction or RBW.
Limit: Vrms_BW < X
3) Spur / ripple mask
Classify narrowband peaks separately; do not fold them into Vrms.
Limit: max spur < X
4) Environment sensitivity
Block airflow and fix cables; repeat measurements and verify stability.
Limit: Δmetric < X across trials
5) Repeatability / reproducibility
Repeat on the same unit and across units; keep BW/ENBW and Tobs identical.
Limit: range or σ < X

Production screens (short-time + automatable + traceable)

Production cannot afford long observation for low-frequency peak-to-peak. Use short-time RMS and spur detection with a fixed bandwidth definition, plus a temperature audit strategy per lot or shift.

Screen 1: short-time RMS
Fixed BW/ENBW and window; catches broadband shifts quickly.
Limit: Vrms_short < X
Screen 2: spur bins
Scan defined bins for fixed peaks (hum/ripple-like behavior).
Limit: spur bin < X
Screen 3: temperature audit
Audit per lot/shift; verifies trend stability across temperature points.
Limit: ΔVrms_temp < X

Reference BOM (starting points only)

These part numbers are provided to speed up fixture and measurement-chain setup. Final selection depends on availability, packaging, and the required bandwidth/precision definition.

Shielding enclosure
Hammond 1590 (diecast aluminum enclosure)
Low-noise power
Analog Devices LT3042 (ultralow-noise LDO)
Low-coupling cable
Belden 9222 (triax cable)
Noise capture ADC
Analog Devices AD7768-1 (precision ΣΔ ADC)
Checklist + data schema (Metric / Condition / Limit / Record)

Use the left checklist to enforce consistent definitions, and the right table to standardize recording fields across design, verification, and production.

Noise Engineering Checklist and Recording Schema Left panel shows design, verification, and production checklists with checkbox items. Right panel shows a mini table with Metric, Condition, Limit, and Record columns. Noise-only checklist (budget → verify → screen) Checklist Design en(f) curve BW / ENBW gain margin Verify 0.1–10Hz p-p Vrms (BW) spur mask repeat floor plan Production short RMS spur bins Record schema Metric Condition Limit (X) Record 0.1–10Hz p-p Tobs X SN/lot Vrms (BW) ENBW X FFT key max spur mask X bin list Vrms_short window X SN/shift Noise-only checklist: keep BW/ENBW + Tobs + floor in every record

Request a Quote

Accepted Formats

pdf, csv, xls, xlsx, zip

Attachment

Drag & drop files here or use the button below.

FAQs: Noise Metrics (0.1–10 Hz & Wideband) — Practical Measurement and Mapping

These FAQs close common long-tail questions about noise specifications, measurement definitions, and mapping noise into resolution. Each answer follows a fixed 4-line, data-oriented structure.

What’s the practical difference between 0.1–10 Hz noise and offset drift?
Likely cause: 0.1–10 Hz noise is random low-frequency fluctuation within a defined band, while drift is a systematic trend that contaminates low-frequency metrics when conditions are not stable.
Quick check: Apply the same 0.1–10 Hz band-pass, then compare “detrended” vs “raw” logs and repeat runs after thermal settling.
Fix: Report low-frequency noise only with fixed Tobs and stabilized environment; treat long-term trend separately in accuracy reporting.
Pass criteria: With fixed BW and Tobs, repeated 0.1–10 Hz results stay within X across trials and detrending does not change the conclusion.
Why does my measured 0.1–10 Hz p-p change when I log longer?
Likely cause: Peak-to-peak is a “max–min” statistic; longer observation increases the chance of catching larger excursions, especially for low-frequency processes.
Quick check: Re-run with multiple fixed Tobs values (e.g., T1, 2×T1, 4×T1) using the same band-pass and compare p-p scaling.
Fix: Freeze Tobs (and the 0.1–10 Hz definition) for reporting; use RMS or percentile metrics only if the window and method are fixed.
Pass criteria: For a fixed Tobs and method, p-p variation across repeated runs stays within X; changes are explainable only by deliberate Tobs changes.
Can I convert 0.1–10 Hz p-p to RMS? What assumptions are required?
Likely cause: A fixed p-p↔RMS factor exists only under stable statistics (near-Gaussian, stationary) and a fixed measurement definition (BW + Tobs + processing).
Quick check: From the same captured record, compute both RMS (within 0.1–10 Hz) and p-p (same window) and derive k = Vpp/Vrms.
Fix: Calibrate k in-system with the chosen BW and Tobs; reuse k only when the definition and environment remain unchanged.
Pass criteria: k stays within X across repeated runs and does not drift with small changes in Tobs; otherwise do not convert.
Why does the datasheet show great noise but my PCB looks worse?
Likely cause: The measurement definition is not matched (BW/ENBW, Tobs, input condition), or the result is dominated by instrument floor, spurs (hum/ripple), or environment sensitivity.
Quick check: Measure chain floor with input short, run an FFT spur scan, and repeat with fixed cabling and a shielded/thermally stable setup.
Fix: Lock the reporting fields (BW/ENBW + Tobs + processing), separate spurs from Vrms, and ensure DUT noise is above chain floor by a margin.
Pass criteria: With matched definitions, the measured noise converges within X of expectation and becomes insensitive to cable touch/airflow within X.
How do I integrate noise density into RMS correctly (ENBW vs BW)?
Likely cause: Using -3 dB bandwidth as “noise BW” or ignoring filter shape causes incorrect Vrms; the correct result depends on ENBW (or integrating en(f) through |H(f)|).
Quick check: Compute ENBW for the actual filter/FFT window and compare Vrms = en·sqrt(ENBW) (white region) against a direct integration estimate.
Fix: Report ENBW explicitly (or the exact integration method) and keep the same window correction when comparing measurements.
Pass criteria: Vrms computed with ENBW matches measured Vrms within X under the same BW/ENBW and processing definition.
What bandwidth should I use to claim a “resolution” number?
Likely cause: “Resolution” is not a single number; it is tied to the measurement bandwidth (or ENBW) and the time window used for averaging/logging.
Quick check: Recompute Δmin (or LSB-equivalent) for two bandwidths and verify the expected trend: BW↑ → Vrms↑ → resolution↓.
Fix: Publish resolution only as “resolution @ BW/ENBW and window”; keep the definition identical for comparisons and procurement specs.
Pass criteria: The claimed resolution remains stable within X when re-tested under the same BW/ENBW and window definition.
Why do chopper INAs show ripple/spurs and how do I budget them?
Likely cause: Chopper/auto-zero action introduces periodic components (ripple/spurs) and possible folding artifacts that appear as narrowband peaks near chopping-related frequencies.
Quick check: Run FFT with a defined RBW/ENBW and confirm fixed-frequency peaks that persist across runs and move with known mode/config changes.
Fix: Treat ripple as spurs (mask/budget separately), and choose measurement BW/window that either excludes the spur or verifies it is below the allowed mask.
Pass criteria: Chopper-related spurs remain below X relative to the floor or outside the stated measurement band by definition.
Why does adding a low-pass filter sometimes make p-p look worse?
Likely cause: Removing higher-frequency content can make slow components (very-low-frequency wander, environment sensitivity) dominate the time record, inflating peak-to-peak within the same Tobs.
Quick check: Apply the same 0.1–10 Hz band-pass definition before comparing p-p, and compare with/without detrending under identical Tobs.
Fix: Use p-p only with a fixed band-pass and stabilized environment; treat out-of-band slow components as a separate stability issue.
Pass criteria: With fixed band-pass and Tobs, p-p decreases or stays within X when LPF is added; any increase is explained by a documented slow component.
How do I separate mains hum from true broadband noise quickly?
Likely cause: Mains hum appears as narrowband peaks at 50/60 Hz and harmonics; broadband noise raises the whole floor rather than forming discrete lines.
Quick check: Run FFT with a fixed RBW/ENBW and look for stable peaks at 50/60 Hz (and multiples) that do not smear across frequency.
Fix: Treat hum as spurs: define a spur mask, then improve shielding/cabling discipline until the peaks fall below the mask.
Pass criteria: Hum peaks are below X relative to the noise floor (or below the defined spur mask) under the fixed measurement setup.
How do I verify instrument noise floor isn’t dominating the result?
Likely cause: If the ADC/FFT chain floor is close to the DUT output noise, the reported number is a chain artifact, not the DUT performance.
Quick check: Measure with input short (chain floor), then measure DUT under the same BW/ENBW; compare the gap and validate scaling with gain changes.
Fix: Reduce BW/ENBW, increase gain (within linear range), or use a lower-floor capture chain so DUT noise exceeds floor by a clear margin.
Pass criteria: DUT noise is higher than chain floor by X (margin), and the measured value follows the expected trend when gain or BW is changed deliberately.
Should I measure noise at the output or refer it to the input?
Likely cause: Output noise depends on gain and output loading; input-referred noise enables fair comparison across gains and front-end configurations.
Quick check: Measure output Vrms under a fixed BW/ENBW, then compute input-referred Vrms_in = Vrms_out / (signal gain) and compare across gains.
Fix: Report both when needed, but always include input-referred noise for comparison and budgeting; keep the gain definition explicit.
Pass criteria: Input-referred results remain consistent within X across different gains under the same measurement definition (BW/ENBW and processing).
My noise spectrum changes when I touch the probe/cable—what does it imply?
Likely cause: Touching changes coupling capacitance and shield reference, injecting mains/RFI or altering leakage paths; the measurement is sensitive to cabling/grounding, not just DUT noise.
Quick check: Fix cable position, use a shielded enclosure, repeat FFT, and compare against a baseline with input short to isolate chain coupling.
Fix: Enforce mechanical cable discipline, shield the DUT area, minimize loop area, and validate with the same BW/ENBW and spur mask before reporting Vrms.
Pass criteria: Touching/moving the cable does not change the reported noise or dominant spur by more than X under the fixed setup.
icnavigator

ICNavigator helps engineers find verified ICs, alternatives, and fast quotes.

Explore
Categories
  • Power Management
  • MCU
  • Automotive ICs
  • Sensors
  • Interface & Transceivers
  • Analog Front-End
Get in Touch
  • Email: hello@icnavigator.com
  • WeChat/WhatsApp: +86-xxxx

© 2025 ICNavigator. All rights reserved.